• 4 Posts
  • 25 Comments
Joined 6 months ago
cake
Cake day: April 4th, 2024

help-circle





  • Depends on what you do with it. Synthetic data seems to be really powerful if it’s human controlled and well built. Stuff like tiny stories (simple llm-generated stories that only use the complexity of a 3-year olds vocabulary) can be used to make tiny language models produce sensible English output. My favourite newer example is the base data for AlphaProof (llm-generated translations of proofs in Math-Papers to the proof-validation system LEAN) to teach an LLM the basic structure of Mathematics proofs. The validation in LEAN itself can be used to only keep high-quality (i.e. correct) proofs. Since AlphaProof is basically a reinforcement learning routine that uses an llm to generate good ideas for proof steps to reduce the size of the space of proof steps, applying it yields new correct proofs that can be used to further improve its internal training data.











  • Public transit seemed relatively safe to me, when I was there 2 years ago. The Amalfi coast is no that great though. Rather go to Naples, use a day for small boat trip to either Capri or Ischia, one day to take the commuter train to Herculaneum. After that take the train from Naples to Salerno, which is a good bit less touristy and ceaper than the real Amalfi coast. Perhaps take a boat from Salerno to look at the Amalfi coast from the sea (more beautiful that way anyway) or enjoye some of the beaches that are reachable with public transit nearby. In Salerno you can also take a train or Bus to Paestum, which is a UNESCO world heritage site has some greek temples from 600 BC.